The Latest Surveys: Here We Go Again!!

Dr. Prospero E. de Vera

The latest Pulse Asia, Manila Standard, and Business World-SWS surveys have generated a new round of analysis, and counter analysis, from supporters and spokesmen of the presidential candidates.

Last night, Tina Monzon-Palma invited me to her show Talk Back to analyze the latest Pulse Asia survey results. I was joined by Pulse Asia President Ronnie Holmes, Aquino campaign spokesman Edwin Lacierda, Gibo Teodoro media head Mike Toledo, and the venerable Ernie Maceda who is Erap Estrada's campaign manager.

The Pulse Asia February 21-25 survey (http://www.pulseasia.com.ph/pulseasia/story.asp?ID=710) declared that Aquino has reclaimed the lead over Villar by a 7 percent margin (36%-29%), that Estrada gained 6% points (12%-18%), and Gibo Teodoro ratings marginally increased to 7% from the January 2010 survey.

Pulse Asia in its media release also stated that the survey was conducted during a period dominated by the following: (1) speech delivered by Senator Manuel B. Villar in the Senate reiterating his innocence in connection with the C-5 road extension project controversy; (2) Senator Panfilo M. Lacson’s departure from the country; (3) official start of the campaign period for national positions; (4) COMELEC’s decision to unseat Pampanga Governor Ed Panlilio; (5) declaration of Lakas-Kampi CMD to field President Gloria Macapagal-Arroyo as its bet for House Speaker in the 15th Congress; (6) arrest of the Morong 43; (7) onset of the El Niño phenomenon; (8) increase in power rates and fluctuations in oil prices; and (9) occurrence of rotating brownouts in different parts of the country and the proposal to grant President Arroyo emergency powers to deal with the energy crisis.

The central questions asked in the show were - Are surveys helping you choose the next president? What and how do the strategists of the candidates use the survey results for their campaign?

While I didn't expect any new earth shaking analysis from Prof. Ronnie Holmes (the Pulse website contains enough analysis and tables), I did expect some new spins from the spokesmen of Aquino, Teodoro, or Estrada.

I was pleasantly surprised by Mike Toledo's grasp of the historical mis-steps by survey groups here and in the US, and his questions on how data gathering can be done in war torn areas.

Manong Ernie was at his best defending Estrada from the "too old" tag given by television viewers, Erap's "Mindanao message", and his sniping of both Aquino and Villar's "saturation" point.

What startled me was the spin given by the Aquino camp when asked about the decrease in Villar's support level. Consistent with his previous media statements, Edwin Lacierda declared that the decrease in Villar's ratings was due to his failure to answer the C-5 corruption allegations. According to Lacierda, the end-January survey did not fully capture the public sentiment about the issue but this new survey has clearly shown that the controversy has affected Villar's ratings.

As a UP Professor who has taught research methods in the graduate school over the past three decades I was shocked by this blatantly misleading analysis and twisting of facts.

If any of my students makes an incorrect interpretation of data he/she would definitely flunk my class!

During the Talk Back show, I had to categorically assert that:

1. Survey groups routinely describe the events happening when a survey is conducted to give the context under which voters preference is generated;

2. But it is impossible to do a one-on-one correlation between certain events (such as the C-5 controversy) and voter's presidential preference (choosing Villar or not). Pulse Asia never claims that there was a correlation. Unfortunately, some campaign spokesmen, and sometimes even the media, make this wrong correlation; and

3. Any such correlation is nothing more than media spin made by supporters of candidates. It is not supported by any empirical or statistical evidence.

If you want to test such a correlation, you have to ask each of the 1,800 survey respondents three questions:

First, are they aware of the C-5 controversy?

Second, among those who are aware, do they believe that the C-5 project is graft ridden? and do they believe Villar is guilty of the allegations of graft and corruption in his sponsorship of the C-5 project?

And third, for those who believe that Villar is guilty, will they vote for him as President if elections were held now?

Then you have to ask the same series of questions for all the other ten events mentioned by Pulse Asia in order to separate the C-5 event vis-a-vis all the other events occurring during the survey period.

Any freshman student of Statistics 101 would know this.

The most recent Manila Standard survey (http://www.manilastandardtoday.com/insideNews.htm?f=2010/march/8/news1.isx&d=2010/march/8) of 2,500 registered voters, and the Business World-SWS survey done during the same survey period says that Villar and Aquino are statistically tied!!

C-5 controvery? Mindanao message? It will be interesting to see how the campaign spokesmen explain or spin this!!